The iPhone has been a game-changer for the mobile world and the software industry in general. I expect the iPhone to continue its dominance of the smartphone for several more years and I think that the success of the iPhone is (in great part) due to the app store. The app store is great for consumers because they have the opportunity to download apps very easily and users have benefitted a lot from the apps. As a user myself, I use a number of apps regularly. The app store has also opened up the mobile app market to developers and eliminated the impact of mobile-carrier walled gardens. This has resulted in a lot of innovation and created value for developers as well as users.
This week’s changes to the app store (and itunesConnect) process for developers prompted me to write about the app store’s evolution over the past 16 months.
The app store has created a market where users are willing to pay money to download mobile apps and (even though most apps make very little money) the app store has created a great opportunity for developers. However, the itunesConnect and app-store process hasn’t been very developer-friendly. Here is a list of some changes that have improved the process over the lifespan of the app store so far.
1. In Sep/Oct 2008, Apple prevented non-users from writing reviews for apps. Until then, many apps got angry 1-star reviews from users who didn’t even bother downloading the app. Limiting reviews to people who actually downloaded the app was a great improvement to the app store
2. In November 2008, Apple introduced the “top paid” app list for each app store category. This was another great improvement. Given the vast amount of junk in the app store, it is hard for good apps to stand above the fray. In this context, the top-paid app list (within each category) has done a good job of improving the visibility of many apps. As examples, our iPhone Government Spending app (released more than six months ago) and our Economy app are both in the category-specific “top-paid apps” list and have benefitted from this over several months since their release dates. This is not to say that other apps (i.e. apps that aren’t in the ‘top paid’ list) are bad. Our Global Warming Personal Calculator app is a good app, but it is not in the top 100 list for its category.
3. For several months last year, some unscrupulous developers exploited a bug that let them update the “release date” of their app on a regular basis. Since apps were sorted on release-date, these developers were able to game the system to push their apps to the top of the list and get more visibility easily. Apple fixed this “release date” logic bug in December, 2008.
4. In June 2009, Apple released iPhone OS 3.0 which supports in-app purchase. From a dev/business perspective, this is one of the most interesting developments in the past year. The V2 version of Economy uses in-app-purchase and I’ll write more about in-app-purchase in a later post.
5. In September 2009, Apple introduced a “Top Grossing” list in addition to the “top paid” list. The latter uses rankings based on the number of users/downloads, while the former is based on net revenues. In recent weeks, the “top grossing” list shows that only around 15% of the top 100 grossing apps are $0.99 apps. This list can encourage developers to charge higher (and more fair) prices for their apps and users get more visibility to these apps through the top-grossing list.
6. In October 2009, Apple introduced freemium support (i.e. support for in-app purchase in free iPhone apps). This was an important development, but I don’t think it was as big as many others made it to be. My analysis of the new freemium policy is here.
7. Until this week, the app-store “release date” category sorted apps by the “release date” of the last version of the app. Some developers gamed the system by releasing frequent updates (without any significant new work) and these updates pushed their apps to the top of the list. That won’t be possible anymore. With this week’s change, apps will be sorted based on their first “release” date and updates will not push an app to the top of the list. This is bound to reduce the number of frivolous app-updates and that may reduce the load on Apple’s app-reviewers. One potential problem is that some developers may re-release the same (old) app as a new app.
8. This week, Apple added a little bit of transparency to the app-review process. In the past, developers had no idea of how far along their app was in the review process, but Apple has now added a new “Status History” page for app submissions. I could see this with my submission of the “Economy V2.0” update. Yesterday morning, my submission of “Economy V2” was moved from the “Waiting for Review” status to the “In review” status by user “Apple” and I could see that in the status-history page. This is a small, but very-good improvement and the added transparency will help developers.
Note that I haven’t attempted to address one of the most controversial aspects of the app-store-review process – the rejection of apps like Google Voice etc. – However, I think it will be a mistake for the government (the FCC) to force Apple to carry any apps. I also hope that Apple works better at clearly communicating the criteria that they use for rejecting apps.
On another complaint, I agree that good apps are sometimes buried in a deluge of bad apps and that it isn’t easy to discover all good apps in the app store. However, I don’t have any specific solution for Apple on this matter.
As an aside , it looks like this weeks itunes update broke the ranking system for paid apps. On Wednesday, Nov 4 (around 4:30pm PST – Seattle time -) , I noticed that the top 100 paid app rankings had gone through a dramatic transformation. The (previously) top-ranked apps were missing and a whole lot of unknown apps were now ranked in the top 100. The problems were fixed shortly before 6pm, but the pattern was repeated again Thursday morning. After that, it looks like Apple restored older (correct) rankings and then froze the rankings until Friday in order to eliminate distortions created by the ranking-system-bugs. Since then, the system seems to be functioning correctly again.
So what more can Apple do ? I think that they should continue making improvements to increase the transparency of the whole review process. This week’s changes are a good step in that direction.
At last year’s (Apple) tech-talk event in Seattle, I had asked the Apple guys to give developers the chance to respond to user review-comments in the app store. Often users ask questions or make feature suggestions and it will be great if developers get a chance to respond to these questions and suggestions. Some reviews tend to be misleading or just outright false and it will be good if developers get a chance to comment on these reviews as well. At the moment, the review section of the app-store is a one-sided dialogue and I think that the app store will improve a lot and users will be better served if Apple gives developers the chance to respond to app store user-review comments.
Apple should do also do a better job with the “release date” field for apps. Currently, by default, the release-date for apps is set as the date on which the app was submitted by the developer (to the app store). If developers set a later release date and the app is approved earlier, the “approval” date is set as the release date. Instead, Apple should just set the release date as the date on which the app is actually released to the app store.
Overall, I’m happy that Apple has been making several improvements to the app store and it will be great if they continue making more improvements such as the ones I’ve mentioned above.
Cascade Software Corporation 
Another improvement they’ve made is that ratings and reviews for the current version of an app are separated out from those for all versions.
Thanks for the note. Yes, this was an important change and many developers had been asking for this change. However, I wouldn’t describe the change as an improvement
The change increased the chances of skewed ratings for apps. It can hurt developers who come out with good V1 releases and it makes it easier for other developers to game the ratings.
As an example of the former, the v1 version of the Economy app had more than 120 ratings and (by a huge margin) it had a plurality of 5-star ratings. However, the first five ratings of the v2 version didn’t include any 5-star ratings and they included (for some inexplicable reason) a couple of 1-star ratings.
The app store now ignores the previous 120+ ratings of the v1 version. It calculates the rating of the app based on the small sample-size of five ratings received by the v2 version. This results in a misleading/skewed average rating for the app.
I think that Apple should avoid displaying a skewed rating by giving equal prominence to the average rating of the current version and the average rating of the app across all versions.